Principal component analysis of Landsat data obtained over the estuary area of the Haihe River in the northwest of the Bohai Bay was performed. The first principal component of Landsat data represents the spectral signature of turbid sea water. The second principal component represents the spectral effect of urban sewage in sea water. The first principal component image of Landsat data received on 4 September 1979 shows the different turbidities of the sea water and the dilution and diffusion of urban sewage in the sea in the rainy season. The second principal component image

of Landsat data received on 19 May 1979 is used to show the different polluted gradations of the sea water in the dry season.The water quality evaluation maps assessed with the multifactor comprehensive index of the organic pollution data measured in situ are used to compare with the principal component images and to calibrate the different polluted gradations obtained from remotely sensed data.